Comparisons of Machine Learning Methods of Statistical Downscaling Method: Case Studies of Daily Climate Anomalies in Thailand

Support Vector ◽

Seasonal Temperature ◽

Statistical Downscaling Method ◽

Downscaling Method

The climate change which is essential for daily life and especially agriculture has been forecasted by global climate models (GCMs) in the past few years. Statistical downscaling method (SD) has been used to improve the GCMs and enables the projection of local climate. Many pieces of research have studied climate change in case of individually seasonal temperature and precipitation for simulation; however, regional difference has not been included in the calculation. In this research, four fundamental SDs, linear regression (LR), Gaussian process (GP), support vector machine (SVM) and deep learning (DL), are studied for daily maximum temperature (TMAX), daily minimum temperature (TMIN), and precipitation (PRCP) based on the statistical relationship between the larger-scale climate predictors and predictands in Thailand. Additionally, the data sets of climate variables from over 45 weather stations overall in Thailand are used to calculate in this calculation. The statistical analysis of two performance criteria (correlation and root mean square error (RMSE)) shows that the DL provides the best performance for simulation. The TMAX and TMIN were calculated and gave a similar trend for all models. PRCP results found that in the North and South are adequate and poor performance due to high and low precipitation, respectively. We illustrate that DL is one of the suitable models for the climate change problem.

Coupling a Markov Chain and Support Vector Machine for At-Site Downscaling of Daily Precipitation

Journal of Hydrometeorology ◽

10.1175/jhm-d-16-0130.1 ◽

2017 ◽

Vol 18 (9) ◽

pp. 2385-2406 ◽

Cited By ~ 5

Author(s):

Yu-Kun Hou ◽

Hua Chen ◽

Chong-Yu Xu ◽

Jie Chen ◽

Sheng-Lian Guo

Keyword(s):

Climate Change ◽

Markov Chain ◽

Statistical Downscaling ◽

Daily Precipitation ◽

Global Climate Models ◽

Support Vector ◽

Statistical Downscaling Method ◽

Calibration And Validation ◽

Downscaling Method ◽

Calibration Methods

Abstract Statistical downscaling is useful for managing scale and resolution problems in outputs from global climate models (GCMs) for climate change impact studies. To improve downscaling of precipitation occurrence, this study proposes a revised regression-based statistical downscaling method that couples a support vector classifier (SVC) and first-order two-state Markov chain to generate the occurrence and a support vector regression (SVR) to simulate the amount. The proposed method is compared to the Statistical Downscaling Model (SDSM) for reproducing the temporal and quantitative distribution of observed precipitation using 10 meteorological indicators. Two types of calibration and validation methods were compared. The first method used sequential split sampling of calibration and validation periods, while the second used odd years for calibration and even years for validation. The proposed coupled approach outperformed the other methods in downscaling daily precipitation in all study periods using both calibration methods. Using odd years for calibration and even years for validation can reduce the influence of possible climate change–induced nonstationary data series. The study shows that it is necessary to combine different types of precipitation state classifiers with a method of regression or distribution to improve the performance of traditional statistical downscaling. These methods were applied to simulate future precipitation change from 2031 to 2100 with the CMIP5 climate variables. The results indicated increasing tendencies in both mean and maximum future precipitation predicted using all the downscaling methods evaluated. However, the proposed method is an at-site statistical downscaling method, and therefore this method will need to be modified for extension into a multisite domain.

Observed Trends in Indices of Daily Temperature Extremes in South America 1960–2000

Journal of Climate ◽

10.1175/jcli3589.1 ◽

2005 ◽

Vol 18 (23) ◽

pp. 5011-5023 ◽

Cited By ~ 294

Author(s):

L. A. Vincent ◽

T. C. Peterson ◽

V. R. Barros ◽

M. B. Marino ◽

M. Rusticucci ◽

...

Keyword(s):

Climate Change ◽

South America ◽

Climate Extremes ◽

Daily Temperature ◽

Maximum Temperature ◽

Daily Maximum ◽

Temperature Extremes ◽

Increasing Trends

Abstract A workshop on enhancing climate change indices in South America was held in Maceió, Brazil, in August 2004. Scientists from eight southern countries brought daily climatological data from their region for a meticulous assessment of data quality and homogeneity, and for the preparation of climate change indices that can be used for analyses of changes in climate extremes. This study presents an examination of the trends over 1960–2000 in the indices of daily temperature extremes. The results indicate no consistent changes in the indices based on daily maximum temperature while significant trends were found in the indices based on daily minimum temperature. Significant increasing trends in the percentage of warm nights and decreasing trends in the percentage of cold nights were observed at many stations. It seems that this warming is mostly due to more warm nights and fewer cold nights during the summer (December–February) and fall (March–May). The stations with significant trends appear to be located closer to the west and east coasts of South America.

Assessing the contribution of multiple forcings to changes in temperature extremes 1981–2020 using CMIP6 climate models

10.5194/egusphere-egu21-7137 ◽

2021 ◽

Author(s):

Mastawesha Misganaw Engdaw ◽

Andrew Ballinger ◽

Gabriele Hegerl ◽

Andrea Steiner

Keyword(s):

Climate Change ◽

Greenhouse Gases ◽

Climate Models ◽

Temporal Trends ◽

Maximum Temperature ◽

Daily Maximum ◽

Temperature Extremes ◽

Cold Temperatures

In this study, we aim at quantifying the contribution of different forcings to changes in temperature extremes over 1981&#8211;2020 using CMIP6 climate model simulations. We first assess the changes in extreme hot and cold temperatures defined as days below 10% and above 90% of daily minimum temperature (TN10 and TN90) and daily maximum temperature (TX10 and TX90). We compute the change in percentage of extreme days per season for October-March (ONDJFM) and April-September (AMJJAS). Spatial and temporal trends are quantified using multi-model mean of all-forcings simulations. The same indices will be computed from aerosols-, greenhouse gases- and natural-only forcing simulations. The trends estimated from all-forcings simulations are then attributed to different forcings (aerosols-, greenhouse gases-, and natural-only) by considering uncertainties not only in amplitude but also in response patterns of climate models. The new statistical approach to climate change detection and attribution method by Ribes et al. (2017) is used to quantify the contribution of human-induced climate change. Preliminary results of the attribution analysis show that anthropogenic climate change has the largest contribution to the changes in temperature extremes in different regions of the world.Keywords: climate change, temperature, extreme events, attribution, CMIP6&#160;Acknowledgement: This work was funded by the Austrian Science Fund (FWF) under Research Grant W1256 (Doctoral Programme Climate Change: Uncertainties, Thresholds and Coping Strategies)

Indices for daily temperature and precipitation based on quality controlled and homogenized data in Madagascar

10.5194/egusphere-egu2020-19970 ◽

2020 ◽

Author(s):

Luc Yannick Andréas Randriamarolaza ◽

Enric Aguilar ◽

Oleg Skrynyk

Keyword(s):

Climate Change ◽

Missing Data ◽

Extreme Temperature ◽

Maximum Temperature ◽

Climate Indices ◽

Low Density ◽

Climate Data ◽

Temperature And Precipitation

Madagascar is an Island in Western Indian Ocean Region. It is mainly exposed to the easterly trade winds and has a rugged topography, which promote different local climates and biodiversity. Climate change inflicts a challenge on Madagascar socio-economic activities. However, Madagascar has low density station and sparse networks on observational weather stations to detect changes in climate. On average, one station covers more than 20 000 km2 and closer neighbor stations are less correlated. Previous studies have demonstrated the changes on Madagascar climate, but this paper contributes and enhances the approach to assess the quality control and homogeneity of Madagascar daily climate data before developing climate indices over 1950 &#8211; 2018 on 28 synoptic stations. Daily climate data of minimum and maximum temperature and precipitation are exploited.Firstly, the quality of daily climate data is controlled by INQC developed and maintained by Center for Climate Change (C3) of Rovira i Virgili University, Spain. It ascertains and improves error detections by using six flag categories. Most errors detected are due to digitalization and measurement.Secondly, daily quality controlled data are homogenized by using CLIMATOL. It uses relative homogenization methods, chooses candidate reference series automatically and infills the missing data in the original data. It has ability to manage low density stations and low inter-station correlations and is tolerable for missing data. Monthly break points are detected by CLIMATOL and used to split daily climate data to be homogenized.Finally, climate indices are calculated by using CLIMIND package which is developed by INDECIS* project. Compared to previous works done, data period is updated to 10 years before and after and 15 new climate indices mostly related to extremes are computed. On temperature, significant increasing and decreasing decade trends of day-to-day and extreme temperature ranges are important in western and eastern areas respectively. On average decade trends of temperature extremes, significant increasing of daily minimum temperature is greater than daily maximum temperature. Many stations indicate significant decreasing in very cold nights than significant increasing in very warm days. Their trends are almost 1 day per decade over 1950 &#8211; 2018. Warming is mainly felt during nighttime and daytime in Oriental and Occidental parts respectively. In contrast, central uplands are warming all the time but tropical nights do not appear yet. On rainfall, no major significant findings are found but intense precipitation might be possible at central uplands due to shortening of longest wet period and occurrence of heavy precipitation. However, no influence detected on total precipitation which is still decreasing over 1950 - 2018. Future works focus on merging of relative homogenization methodologies to ameliorate the results.-------------------*INDECIS is a part of ERA4CS, an ERA-NET initiated by JPI Climate, and funded by FORMAS (SE), DLR (DE), BMWFW (AT), IFD (DK), MINECO (ES), ANR (FR) with co-funding by the European Union (Grant 690462).

Application of Artificial Neural Networks to Project Reference Evapotranspiration under Climate Change Scenarios

10.21203/rs.3.rs-603225/v1 ◽

2021 ◽

Author(s):

Junaid Maqsood ◽

Aitazaz A. Farooque ◽

Farhat Abbas ◽

Travis Esau ◽

Xander Wang ◽

...

Keyword(s):

Climate Change ◽

Neural Networks ◽

Artificial Neural Networks ◽

General Circulation ◽

Reference Evapotranspiration ◽

Maximum Temperature ◽

Climate Change Scenarios ◽

Artificial Neural

Abstract Evapotranspiration, one of the major elements of the water cycle, is sensitive to climate change. The main objective of this study was to examine the response of reference evapotranspiration (ET0) under various climate change scenarios using artificial neural networks and a general circulation model (GCM) - the Canadian Earth System Model Second Generation (CanESM2). The Hargreaves method was used to calculate ET0 for western, central, and eastern parts of Prince Edward Island. The two input parameters of the Hargreaves method; daily maximum temperature (Tmax), and daily minimum temperature (Tmin) were projected using CanESM2. The Tmax and Tmin were downscaled with the help of statistical downscaling and simulation model (SDSM) for three future periods 2020s (2011–2040), 2050s (2041–2070), and 2080s (2071–2100) under three representative concentration pathways (RCP’s) including RCP 2.6, RCP P4.5, and RCP 8.5, and the. Temporally, there were major changes in Tmax, Tmin, and ET0 for the 2080s under RCP8.5. The temporal variations in ET0 for all RCPs matched the reports in the literature for other similar locations and for RCP8.5 it ranged from 1.63 (2020s) to 2.29 mm/day (2080s). As a next step, a one-dimensional convolutional neural network (1D-CNN), long-short term memory (LSTM), and multilayer perceptron (MLP) were used for estimating ET0 due to the non-linear behavior of ET0 and the limited meteorological input data. High coefficient of correlation (r > 0.95) values for both calibration and validation periods showed the potential of the artificial neural networks in ET0 estimation. The results of this study will help decision makers and water resource managers to quantify the availability of water in future for the island and to optimize the use of island water resources on a sustainable basis.

Trends in Extreme Climate Indices in Cherrapunji for the period 1979 to 2020

10.21203/rs.3.rs-706445/v1 ◽

2021 ◽

Author(s):

Raju Kalita ◽

Dipangkar Kalita ◽

Atul Saxena

Keyword(s):

Climate Change ◽

Negative Trend ◽

Maximum Temperature ◽

Trend Test ◽

Climate Indices ◽

Daily Maximum ◽

Positive Trend ◽

Extreme Climate

Abstract We have used Mann-Kendall trend test and Sen’s slope estimator method to find out significant changes in extreme climate indices for daily temperature as well as precipitation over the period 1979 to 2020 in Cherrapunji. In the present study, a total of 24 precipitation and temperature based extreme climate indices were calculated using RClimDex v 1.9-3. Among 24 indices, 7 were derived from number of days above nn mm rainfall (Rnn) according to Indian Meteorological Department (IMD) convention and the rest were in accordance with the Expert Team on Climate Change Detection and Indices (ETCCDI). It was observed that, among all the indices, consecutive dry days (CDD), summer days (SU25) and very light rainfall (VLR) days increased significantly with 0.54, 1.58 and 0.14 days/year respectively, while only consecutive wet days (CWD) decreased significantly with 0.36 days/year. A slight negative trend was also observed in case of tropical nights (TR20) and among the other precipitation indices as well. Again, the indices associated with daily maximum temperature increased significantly with annual change of 0.06 to 0.07 ⁰C/year. And for indices associated with daily minimum temperature, almost no change or a slight negative change was observed, except a significant positive trend in February and significant negative trend in November for TNN only. The analysis reveals that some of the extreme climate indices which explains the climatic conditions of Cherrapunji has changed a lot over the period of 42 years and if this trend continues then Cherrapunji will be under threat when it comes to climate change.

Historical Variability And Future Changes In Seasonal Extreme Temperature Over Iran

10.21203/rs.3.rs-739242/v1 ◽

2021 ◽

Author(s):

Azar Zarrin ◽

Abbasali Dadashi-Roudbari ◽

Samira Hassani

Keyword(s):

Climate Change ◽

General Circulation ◽

Extreme Temperature ◽

Maximum Temperature ◽

Cmip5 Models ◽

Important Indicator ◽

Taylor Diagram ◽

Temperature Indices

Abstract The extreme temperature indices (ETI) are an important indicator of climate change, the detection of their changes over the next years can play an important role in the Climate Action Plan (CAP). In this study, four temperature indices (Mean of daily minimum temperature (TN), Mean of daily maximum temperature (TX), Cold-spell duration index (CSDI), and Warm-spell duration index (WSDI)) were defined by ETCCDI and two new indices of the Maximum number of consecutive frost days (CFD) and the Maximum number of consecutive summer days (CSU) were calculated to examine ETIs in Iran under climate change conditions. We used minimum and maximum daily temperature of five General circulation models (GCMs) including HadGEM2-ES, IPSL-CM5A-LR, GFDL-ESM2M, MIROC-ESM-CHEM, and NorESM1-M from the set of CMIP5 Bias-Correction models. We investigated Two Representative Concentration Pathway (RCP) scenarios of RCP4.5 and RCP8.5 – during the historical (1965-2005) and future (2021-2060 and 2061-2100) periods. The performance of each model was evaluated using the Taylor diagram on a seasonal scale. Among models, GFDL-ESM2M and HadGEM2-ES models showed the highest, and NorESM1-M and IPSL-CM5A-LR models showed the lowest performance in Iran. Then an ensemble model was generated using Independence Weighted Mean (IWM) method. The results of multi-model ensembles (MME) showed a higher performance compared to individual CMIP5 models in all seasons. Also, the uncertainty value was significantly reduced, and the correlation value of the MME model reached 0.95 in all seasons. Additionally, it is found that WSDI and CSU indices showed positive anomalies in future periods and CSDI and CFD showed negative anomalies throughout Iran. Also, at the end of the 21st century, no cold spells are projected in almost every part of Iran. The CSU index showed that Iran's summer days are increasing sharply, according to the results of the RCP8.5 scenario in spring (MAM) and autumn (SON), the CSU will increase by 18.79 and 20.51 days, respectively at the end of the 21st century. It is projected that in the future, the spring and autumn seasons will be shorter and, summers, will be much longer than before.

Spatio-temporal variation of extreme indices derived from observed and reanalysis products for detection of climate change across India

10.5194/egusphere-egu2020-5837 ◽

2020 ◽

Author(s):

Sachidanand Kumar ◽

Kironmala Chanda ◽

Srinivas Pasupuleti

Keyword(s):

Climate Change ◽

Extreme Precipitation ◽

Daily Precipitation ◽

Temporal Trends ◽

Maximum Temperature ◽

Daily Maximum ◽

Spatio Temporal ◽

Precipitation And Temperature

AbstractThis article reports the research findings in a recent study (Kumar et al., 2020) that utilizes eight indices of climate change recommended by the Expert Team on Climate Change Detection and Indices (ETCCDI) for analyzing spatio-temporal trends in extreme precipitation and temperature at the daily scale across India. Observed gridded precipitation (1971-2017) and temperature (1971-2013) datasets from India Meteorological Department (IMD) are used along with reanalysis products from Climate Prediction Centre (CPC). The trends are estimated using non-parametric Mann-Kendall (MK) test and regression analysis. The trends in &#8216;wet days&#8217; (daily precipitation greater than 95th percentile) and &#8216;dry days&#8217; (daily precipitation lower than 5th percentile) are examined considering the entire year (annual) as well as monsoon months only (seasonal). At the annual scale, about 13% of the grid locations indicated significant trend (either increasing or decreasing at 5% significance level) in the index R95p (rainfall contribution from extreme &#8216;wet days&#8217;) while 20% of the locations indicated significant trend in R5p (rainfall contribution from extreme &#8216;dry days&#8217;). For the seasonal analysis (June to September), the corresponding figures are nil and 21% respectively. The spatio-temporal trends in &#8216;warm days&#8217; (daily maximum temperature greater than 95th percentile), &#8216;warm nights&#8217; (daily minimum temperature greater than 95th percentile), &#8216;cold days&#8217; (daily maximum temperature lower than 5th percentile) and &#8216;cold nights&#8217; (daily minimum temperature lower than 5th percentile) are also investigated for the aforementioned period. The number of &#8216;warm days&#8217; per year increased significantly at 14% of the locations, while the number of &#8216;cold days&#8217;, &#8216;warm nights&#8217; and &#8216;cold nights&#8217; per year decreased significantly at several (42%, 34% and 39%) of the locations. The extreme temperature indices are also investigated for the future using CanESM2 projected data for RCP8.5 after suitable bias correction. Most of the locations (49% to 84%) indicate significant increasing (decreasing) trend in &#8216;warm days&#8217; (&#8216;cold days&#8217;) in the three epochs, 2006-2040, 2041-2070 and 2071-2100. Moreover, most locations (60% to 81%) show an increasing trend in &#8216;warm nights&#8217; and a decreasing trend in &#8216;cold nights&#8217; in all the epochs. A similar investigation for the historical and future periods using CPC data as the reference indicates that the trends, on comparison with IMD observations, seem to be in agreement for temperature extremes but spatially more extensive in case of CPC precipitation extremes.Keywords: extreme precipitation and temperature, climate change indices, spatio-temporal variation, IndiaReferences:Kumar S., Chanda, K., Srinivas P., (2020), Spatiotemporal analysis of extreme indices derived from daily precipitation and temperature for climate change detection over India, Theoretical and Applied Climatology, Springer, In press, DOI: 10.1007/s00704-020-03088-5.

Evaluation of a statistical downscaling procedure for the estimation of climate change impacts on droughts

Natural Hazards and Earth System Science ◽

10.5194/nhess-9-879-2009 ◽

2009 ◽

Vol 9 (3) ◽

pp. 879-894 ◽

Cited By ~ 38

Author(s):

L. Vasiliades ◽

A. Loukas ◽

G. Patsonas

Keyword(s):

Climate Change ◽

Water Resources ◽

Statistical Downscaling ◽

General Circulation ◽

Standardized Precipitation Index ◽

Circulation Model ◽

Economic Losses ◽

Monthly Precipitation ◽

Statistical Downscaling Method ◽

Downscaling Method

Abstract. Despite uncertainties in future climates, there is considerable evidence that there will be substantial impacts on the environment and human interests. Climate change will affect the hydrology of a region through changes in the timing, amount, and form of precipitation, evaporation and transpiration rates, and soil moisture, which in turn affect also the drought characteristics in a region. Droughts are long-term phenomena affecting large regions causing significant damages both in human lives and economic losses. The most widely used approach in regional climate impact studies is to combine the output of the General Circulation Models (GCMs) with an impact model. The outputs of Global Circulation Model CGCMa2 were applied for two socioeconomic scenarios, namely, SRES A2 and SRES B2 for the assessment of climate change impact on droughts. In this study, a statistical downscaling method has been applied for monthly precipitation. The methodology is based on multiple regression of GCM predictant variables with observed precipitation developed in an earlier paper (Loukas et al., 2008) and the application of a stochastic timeseries model for precipitation residuals simulation (white noise). The methodology was developed for historical period (1960–1990) and validated against observed monthly precipitation for period 1990–2002 in Lake Karla watershed, Thessaly, Greece. The validation indicated the accuracy of the methodology and the uncertainties propagated by the downscaling procedure in the estimation of a meteorological drought index the Standardized Precipitation Index (SPI) at multiple timescales. Subsequently, monthly precipitation and SPI were estimated for two future periods 2020–2050 and 2070–2100. The results of the present study indicate the accuracy, reliability and uncertainty of the statistical downscaling method for the assessment of climate change on hydrological, agricultural and water resources droughts. Results show that climate change will have a major impact on droughts but the uncertainty introduced is quite large and is increasing as SPI timescale increases. Larger timescales of SPI, which, are used to monitor hydrological and water resources droughts, are more sensitive to climate change than smaller timescales, which, are used to monitor meteorological and agricultural droughts. Future drought predictions should be handled with caution and their uncertainty should always be evaluated as results demonstrate.

Comparison of Three Statistical Downscaling Methods and Ensemble Downscaling Method Based on Bayesian Model Averaging in Upper Hanjiang River Basin, China

Advances in Meteorology ◽

10.1155/2016/7463963 ◽

2016 ◽

Vol 2016 ◽

pp. 1-12 ◽

Cited By ~ 5

Author(s):

Jiaming Liu ◽

Di Yuan ◽

Liping Zhang ◽

Xia Zou ◽

Xingyuan Song

Keyword(s):

Climate Change ◽

River Basin ◽

Statistical Downscaling ◽

Bayesian Model ◽

Bayesian Model Averaging ◽

Model Averaging ◽

Support Vector ◽

Simulation Accuracy ◽

Downscaling Method ◽

Hanjiang River

Many downscaling techniques have been developed in the past few years for projection of station-scale hydrological variables from large-scale atmospheric variables to assess the hydrological impacts of climate change. To improve the simulation accuracy of downscaling methods, the Bayesian Model Averaging (BMA) method combined with three statistical downscaling methods, which are support vector machine (SVM), BCC/RCG-Weather Generators (BCC/RCG-WG), and Statistics Downscaling Model (SDSM), is proposed in this study, based on the statistical relationship between the larger scale climate predictors and observed precipitation in upper Hanjiang River Basin (HRB). The statistical analysis of three performance criteria (the Nash-Sutcliffe coefficient of efficiency, the coefficient of correlation, and the relative error) shows that the performance of ensemble downscaling method based on BMA for rainfall is better than that of each single statistical downscaling method. Moreover, the performance for the runoff modelled by the SWAT rainfall-runoff model using the downscaled daily rainfall by four methods is also compared, and the ensemble downscaling method has better simulation accuracy. The ensemble downscaling technology based on BMA can provide scientific basis for the study of runoff response to climate change.